Revista de Biología Marina y Oceanografía Vol. 50, Nº3: 479-489, diciembre 2015 DOI: 10.4067/S0718-19572015000400007 ARTICLE Operational interactions between the South American sea lion Otaria flavescens and purse seine fishing activities in northern Chile Interacciones operacionales entre lobo marino común Otaria flavescens y actividades de pesca de cerco en el norte de Chile
Andrés González1, Rodrigo Vega1 and Eleuterio Yáñez2
1Departamento de Evaluación de Pesquerías, División de Investigación Pesquera Instituto de Fomento Pesquero, Blanco 839, Valparaíso, Chile. [email protected] 2Escuela de Ciencias del Mar, Pontificia Universidad Católica de Valparaíso, P.O. Box 1020, Valparaíso, Chile
Resumen.- Se analizan variables temporales, espaciales, ambientales, ecológicas y pesqueras que contribuyen a la variación del número de lobos marinos comunes (Otaria flavescens) que interactúan con las faenas de pesca de cerco industrial de la flota que opera en el norte de Chile (18°21’-24°00’S). Observadores científicos fueron embarcados en buques cerqueros entre febrero 2010 y diciembre 2011. Los datos fueron modelados usando modelos lineales generalizados (MLG) y modelos lineales generalizados mixtos (MLGM). Los resultados muestran que la variabilidad en el número de lobos marinos atraídos por las operaciones de pesca para alimentarse o intentar hacerlo, se explica por las variables: Hora del lance, Distancia a las colonias de lobos más cercanas, Temperatura superficial del mar, Especie objetivo del lance, Latitud y Número de aves marinas en torno a la embarcación. Luego del análisis detallado de las variables significativas, se puede concluir que los principales mecanismos que estarían actuando sobre la interacción entre lobos marinos y la operación de pesca serían: 1) una estrategia utilizada por los otáridos para disminuir el estrés por calor en el período reproductivo (verano), saliendo al mar en horas de máxima radiación solar (al mediodía), lo que incrementa notablemente su presencia dentro del cerco en lances realizados a menos de 20 millas náuticas de las colonias y 2) durante períodos cálidos, la presencia de ondas Kelvin descendentes disminuiría la disponibilidad de anchoveta (Engraulis ringens) en la zona, lo que provoca que la flota reoriente sus operaciones hacia el jurel (Trachurus murphyi), recurso que generalmente se captura a mayores distancias de la costa y de las colonias, disminuyendo significativamente la interacción con los lobos marinos.
Palabras clave: Lobo marino común, Engraulis ringens, Índice ENSO Multivariado, Pesca de cerco
Abstract.- This study analyzes temporal, spatial, environmental, ecological and fishing variables that contributed to variations in the number of South American sea lions (Otaria flavescens) that interacted with the industrial purse seine fishing operations in the north of Chile (18°21’-24°00’S). Scientific observers were placed onboard purse seiners vessels between February 2010 and December 2011. Data were modeled using generalized linear models (GLM) and generalized linear mixed models (GLMM). The results show that the variations in the number of sea lions attracted to fishing operations for feeding or attempting to feed on catches, depends on the following variables: Time of haul, Distance to the closest sea lion colony, Sea surface temperature, Target species of the haul, Latitude, and Number of seabirds circling the vessel. After detailed analysis of the significant variables, it can be concluded that the main mechanisms involved in the interaction between the sea lions and the fishing operations were: 1) a strategy used by the sea lions to decrease heat stress during the breeding period (summer) by entering the water when solar radiation is at its peak (around midday), thus notably increasing their presence within the nets during sets that are less than 20 nautical miles from the colonies; and 2) during warm periods the presence of downwelling Kelvin waves decreases the availability of anchovies (Engraulis ringens) in the area. This causes the fleet to change target species and focus on jack mackerel (Trachurus murphyi), which is generally captured further from the coast and the colonies, significantly decreasing the interaction with local sea lions.
Key words: South American sea lion, Engraulis ringens, Multivariate ENSO Index, Purse seine
INTRODUCTION In areas where the coast is intensively utilized for and marine mammals (Bjørge et al. 2002). In this context harvesting and farming marine resources, there is a interactions between pinnipeds and fishing operations potential for interactions between fish-related industries have been widely documented in the international
Vol. 50, Nº 3, 2015 479 Revista de Biología Marina y Oceanografía literature (Ainley et al. 1982, Bonner 1982, Northridge 1985, the fishing operation (increased work time), reducing Hückstädt & Antezana 2003, Goetz et al. 2008, Reyes et catches and in some cases damaging fishing gear al. 2013). It is known that such interactions will occur in (Beverton 1985, Wickens 1995, Read 2008). The any place in the world where the distributions between operational interactions can also be negative for marine pinnipeds and fishing activities overlap (Riedman 1990, top predators due to the possibility of the animals coming Wickens 1995). into direct contact with the fishing system, leading to injury and even death (Oliva et al. 2003, Read 2008, Fertl Purse seine fishing operations conducted in the north 2009, Reyes et al. 2013). of Chile are denominated as Fishing Units XV, I and II for Chilean Administrative Regions for anchovy SUBPESCA The effects of the interaction on the pinnipeds (seals, 2011a)1 and jack mackerel (SUBPESCA 2011b)2. In the last sea lions and walrus) can be positive, due to the obvious decade, landings of these fisheries have represented on advantages of preying on a concentrated food source average 23% of fish landings in Chile (SERNAPESCA that is easy to access (Hückstädt & Antezana 2003). The 2002-2011)3. In the Administrative Regions mentioned, fishing operations concentrate the food source; leading both species are caught by the same fleet, which to a reduction in the amount of energy the sea lions spend comprises 66 purse seine fishing vessels, averaging 34 m in feeding (Fertl 2009). 3 of length (range between 22 and 44 m), and 400 m of hold The main operational interactions between sea lions 3 capacity (range between 140 and 670 m ; Bernal et al. and fisheries reported in Chile are predation of catch, 4 2010) . The operations of this fleet cover an extensive bycatch and incidental mortality of sea lions (Hückstädt area from the port of Arica in the north (18°21’S; 70°19’W) & Antezana 2003, Sepúlveda et al. 2007, Reyes et al. 2013), to Antofagasta in the south (23°38’S; 70°24’W), and while the damage to fishing gear is important in artisanal reaching 90 nautical miles from the coast (SUBPESCA fisheries (Sepúlveda et al. 2007). Hückstädt & Antezana 2011a). This area sustains a large population of South (2003) reported an average interaction of 21 sea lions per American sea lions (Otaria flavescens), with population set while Reyes et al. (2013), working with the trawling reaching 59,657 individuals distributed over 44 breeding industrial fleet that operates in the south-central Chile and haul-out colonies, thus representing 43.6% of the indicates a catch rate of 1.2 sea lion/trawl-1 and a mortality total population of this species in Chile (Bartheld et al. of 14.6%. 2008). Bycatch records are used to determine the extent Though sea lions are considered generalist predators, (temporal, spatial, by species) and magnitude (number of located on the upper levels of the food web (Beverton individuals per species) of these events, while studying 1985), in the study area the South American sea lion feeds the interactions, abundance of species that interact with mainly on anchovy (Sielfeld et al. 1997, Sielfeld 1999, fisheries, environmental and operational data are used to Arias-Schreiber 2003), which might lead to its interaction understand the nature of these interactions, and the with the purse seine fishing fleet that targets this species. importance of the factors that influence the level of these interactions. This is important for identifying specific Two types of interaction can be defined between mitigation solutions for the particular fishery (ACAP 2012). marine mammals with fishing activities: a biological (indirect) interaction due to competition for food, and an Considering the importance of understanding the main operational (direct) interaction due to predation, where mechanisms underlying the interaction processes between the animals enter the purse seine nets to feed, occasionally non-target species and fishing operations in the context leading to their capture in the net. In general, operational of an ecosystem approach to fisheries management (FAO interactions negatively affect the fishers by disturbing 2003), the current study aims to analyze different factors
1SUBPESCA. 2011a. Cuota global anual de captura para las unidades de pesquería de anchoveta y sardina española, regiones XV, I y II, año 2012. Informe Técnico (R Pesq), 108 pp. Subsecretaría de Pesca y Acuicultura, Valparaíso. 2SUBPESCA. 2011b. Cuota global anual de captura de jurel, para las unidades de pesquería de la XV-II, III-IV, V-IX y XIV-X regiones, año 2012. Informe Técnico (R Pesq), 129 pp. Subsecretaría de Pesca y Acuicultura, Valparaíso. 3SERNAPESCA. 2002-2011. Anuarios estadísticos de pesca. Servicio Nacional de Pesca, Ministerio de Economía, Fomento y Reconstrucción, Chile.
480 González et al. South American sea lion and purse seine fishing in northern Chile (temporal, environmental, spatial, ecological and fishing) on February 2007 by Bartheld et al. (2008) was used to which may contribute to variations in the number of South calculate size and geographic position of the sea lion American sea lions that interact with the industrial purse colony centers (Fig. 1). A sea lion colony center was seine fleet operating in the north of Chile. defined as a group of colonies within a radius of 10 nautical miles around the biggest colony of the group, where the geographic position of the biggest colony was chosen MATERIALS AND METHODS as the location of the center, and the size of the center was defined as the sum of the individual groups of DATA COLLECTION colonies. The computer software packages PBSmapping, The study area comprises the waters of the Chilean RODBC and R language RColorBrewer (R Development Exclusive Economic Zone from the port of Arica (18°21’S; Core Team 2006) were used to calculate the distance in 70°19’W) to the south of the region of Antofagasta nautical miles (nm) from the position at the start of the set (23°38’S; 70°24’W). The data used were obtained by and the position to the closest colony center. Finally, the scientific observers (SO) from the Instituto de Fomento Multivariate ENSO (El Niño Southern Oscillation) Index Pesquero (IFOP; Fisheries Development Institute), as part (MEI) was also included in the analysis. This index of the North Zone Pelagic Fishing Situation Research describes the atmospheric and oceanic conditions in the Program. Information was collected by observers aboard tropical Pacific Ocean (30°S-30°N), adjusted seasonally vessels from 258 hauls over 166 trips performed between with regard to the reference period 1950-1993 (Wolter & February 2010 and December 2011. This observation effort Timlin 1998)6. Negative values of the standardized MEI (166 trips) represented 1% of the total effort applied by represent cold conditions, while positive values represent the fleet in the study period (Gabriela Böhm, pers. comm.). warm conditions). All the aforementioned factors were All the observations recording by SO were made during input into the model as independent variables. The daylight sets. The SO recorded the date, time, and response variable of the model was defined as the number geographic position of all sets, as well as sea surface of South American sea lion feeding or attempting to feed temperature (SST), target species and catch per species on the catch from each haul (Table 1). (in tons). A proxy of ship size, the storage capacity (ton) was also taken. During each haul, the SO counted the number of sea lions attracted by fishing operations for STATISTICAL ANALYSIS feeding or attempting to feed on catches. The number of Regression analysis was performed using generalized other purse seine fishing vessels operating nearby was linear models (GLM) and generalized linear mixed models also recorded, where distance of observation is depending (GLMM). The response variable, the number of sea lions of sighting conditions (radius between 5 and 10 nautic interacting with fishing operations, was given by miles). In order to count the sea lions, the protocol used counting data that was therefore discrete and non- was developed for the IFOP scientific observer program negative. The GLM and GLMM allowed increased (Bernal et al. 2012)5, which stipulates that the counting flexibility in the assumptions compared with traditional should be conducted on daylight sets, on the winch side regression techniques, allowing direct specification of (starboard), maintaining a fixed counting location at the the error distribution (Gill 2001) and integration of moment the net is pulled onto the vessel, and in a 250 m continuous and categorical data into one single setting. hemisphere centered alongside the vessel (Melvin et al. The GLMM cover the dependence between the 2009, Bernal et al. 2012). The time of the haul was used to observations, incorporating two types of effects: direct estimate the intensity of the solar radiation, assuming effects (fixed effects), which are the explanatory variables; that the level of intensity increases from dawn and reaches and the indirect effects (random effects), which include a a maximum around midday, and then decreasing until set number of observations (hauls). The first represent sunset (Campagna & LeBoeuf 1988). Abundance and the mean or ‘typical’ level of the response variable and geographic positions of sea lions colonies data recorded the second represent the specific deviations by grouped
5Bernal C, J Azócar, A González, L Ossa, V Escobar, M San Martín, JC Saavedra, C Villouta, R Bello & J Castillo. 2012. Programa de Seguimiento de las Principales Pesquerías Nacionales. Informe Final Proyecto: Investigación Programa Observadores Científicos, 2012. Instituto de Fomento Pesquero, Valparaíso, 169 pp. 6
Vol. 50, Nº 3, 2015 481 Revista de Biología Marina y Oceanografía Figure 1. Overlap between the South American sea lion colony centers and the centers of upwelling and fishing activity in the study zone / Sobreposición entre los focos de colonias de lobos marinos comunes, respecto de los focos de surgencias y actividad pesquera en la zona de estudio
Table 1. Temporal, spatial, environmental, ecological and fishing variables included in the analysis / Variables temporales, espaciales, ecológicas y pesqueras incluidas en los análisis
482 González et al. South American sea lion and purse seine fishing in northern Chile level, which are incorporated into the marginal the data allowed the exclusion of some variables, such as distribution (Bates 2010). In addition to the GLM and Longitude and Distance from the set to the coastline, GLMM analyses, generalized additive models (GAM) were since both were highly correlated with the variable also fit in order to allow the exploration of non-linear Distance to closest colony center. functional relations between the response variable and The final model showing the best fit was the GLMM the independent variables. The GAM replaces the linear that included the random effect of Fishing trip (IdViaje), predictor of the GLM with an additive predictor allowing within the random effect Fishing vessel (IdBuque). Due the combination of linear with non-linear and non- to the nested structure of the data, mixed modeling parametric relations (Hastie & Tibshirani 1990). Before techniques were applied to allow the inclusion of adding the independent variables into the model, the correlation between observations of the same grouping absence of co-linearity was checked and the co-variables level (Zuur et al. 2009). In our case, the hauls conducted were put onto the same measurement scale by a process during the same fishing trip are probably more related to of standardization, in order to be able to compare their each other than to hauls from other trips, and fishing respective estimated coefficients. Poisson and negative trips of the same vessel are also more connected to one binomial distributions were used to model the number of another than to trips of other ships. sea lions interacting with fishing operations as a function of the explanatory variables, because they are the best When analyzing the abundance of sea lions interacting discrete probability distributions for describing count with the purse seiners in relation to explanatory variables, data (Crawley 2013). The models were first built with a the model indicated the significant effect of Distance to Poisson distribution. After checking and confirming closest colony center, SST, Latitude, Number of vessels overdispersion, negative binomial distribution was used operating in the area, Number of seabirds and Time of instead (Passadore et al. 2015). To choose the best-fitting haul (Table 2). The number of sea lions interacting with model, explanatory variables were selected manually by the purse seiners tends to decrease when the number of deletion from a complex initial model (backward). Analysis vessels increased (Fig. 2a), and when fishing operations of variance (ANOVA) was used to compare the resulting were performed further from the centers of the colonies model with the previous model of the simplication (Fig. 2b). Moreover, sea lions showed a major level of procedure. When the removal of a variable produced a interactions in the northern part of the study area and significant increase in model’s deviance, that variable was then decrease their number as we move south to finally considered statistically significant and, thus, returned and present a secondary peak in the southern part of the area retained in the model. When the removal produced (Fig. 2c). Higher levels of abundance of sea lions attending nonsignificant change in deviance, the variable was left out of the model (Passadore et al. 2015). Variables were Table 2. Result of the fit of the GLMM model for the number of South American sea lions interacting with purse seine fishing sequentially removed until the model contained only operations in northern Chile / Resultado del ajuste del MLGM para significant terms; that model was considered as the el número de lobos marinos comunes interactuando con las minimum adequate model (Crawley 2013). operaciones de pesca de cerco en el norte de Chile The following statistics packages were used for these analyses: lme4 (GLMM; Bates 2010), mgcv (GAM; Wood 2008), and MASS (GLM; Venables & Ripley 2002), available in R language (R Development Core Team 2011). The Akaike Information Criterion (AIC) was used to select the best model. AIC is a measure of the goodness of fit including parsimony. The model determined by parsimony is defined as a model that fits data and so it includes few parameters needed (Burnham & Anderson 2002).
RESULTS The average number (± S.D.) of sea lions that were seen to interact over a total of 258 fishing sets was 55 (± 56) per haul, in a range of 0 to 290. A preliminary analysis of
Vol. 50, Nº 3, 2015 483 Revista de Biología Marina y Oceanografía Figure 2. Graphs of the GAM fit. a) Number of vessels, b) Distance to closest colony center, c) Latitude, d) SST, e) Number of seabirds, f) Time of haul; S(x) represents the smoothed spline function of the predictor variable / Representaciones gráficas del ajuste del MAG. a) Número de buques, b) Distancia al centro de colonias más cercana, c) Latitud, d) Temperatura superficial del mar, e) Número de aves marinas, f) Hora del lance; S(x) representa la función spline suavizada de la variable predictora indicada
the fishing operations were observed between 16 to 17°C of SST, and tends to decrease when the SST increased (Fig. 2d). Furthermore, the number of sea lions interacting increases according as seabird abundance attending purse seine vessels increases (Fig. 2e). Time of haul showed higher values of sea lion interaction around midday and early afternoon (Fig. 2f). Finally, the level of interaction of sea lions was significantly higher (Wilcoxon rank sum test; P-value < 0.001; Fig. 3) when the fleet was capturing anchovy (mean 59 sea lions per set), regarding fishing operations on jack mackerel (mean 29 sea lions per set). The operation over anchovy was more coastal (mean 20 nm from coast), that over jack mackerel (mean 34 nm from coast). High significance was found in the interaction of the variables Distance from haul to the closest colony center and Time of haul, where an important increase in the number of sea lions was observed around midday during sets that took place less than 20 nm from a colony center Figure 3. Boxplot of the number of sea lions interacting with fishing during the breeding period, from mid-December to mid- differentiated by target species / Box plot del número de lobos March (Fig. 4a). This effect disappears in non- marinos comunes interactuando según especie objetivo reproductive period, where were observed similar levels
484 González et al. South American sea lion and purse seine fishing in northern Chile Figure 4. Graphs of the interaction between the Distance to the closest colony center and Time of haul: a) During breeding season, b) During non-breeding season / Representación gráfica de la interacción entre Distancia al foco de colonias más cercana y Hora del lance. a) Periodo reproductivo, b) Periodo no reproductivo
Figure 5. Graphs of interactions based on the GLMM fit. a) Number of vessels-MEI, b) Target species-MEI / Representaciones gráficas del ajuste del MLGM. a) Número de buques-MEI, b) Especie objetivo-MEI
DISCUSSION in the abundance of sea lions during sets that took place The association of the sea lions with coastal fisheries to less and more than 20 nm from a colony center (Fig. has been recorded previously in virtually all areas where 4b). breeding and non-breeding colonies occur (Bastida et al. 2007). Feeding of marine mammals in association with The individual contribution of the Number of vessels fishing operations is a learned behavior leading to the operating in the area and MEI was not significant since a increase of the number of individuals seeking out fishing large part of its effect is already considered in the systems to find food easily. According to Königson et al. interaction between these variables, which was highly (2006), a prior feeding experience may encourage sea lions significant, showing an escalated effect on the response to return to feeding areas where they have been variable for different levels of MEI (Table 2; Fig. 5a). The successful. It has been suggested that this type of interaction between target species and MEI showed that feeding behavior is transferred from generation to the abundance of sea lions decreased as we move from generation by observation and participation (Fertl & cold conditions (negative values of the standardized MEI), Leatherwood 1997). to warm conditions (positive values of the standardized MEI), for both target species (Fig. 5b). In our study, the average number of sea lions interacting with purse seine operations was 55 per haul, which is more than the figure reported by Hückstädt & Antezana
Vol. 50, Nº 3, 2015 485 Revista de Biología Marina y Oceanografía (2003) for the purse seine fleet targeting jack mackerel in One of the main sources of variation in the Humboldt the VIII region of Chile (36°00´S-38°30´S), which reported Current System derives from coastally trapped Kelvin an average of 21 individuals per haul (range: 0-50). This Waves (KW) which are originates by wind variations in difference may be explained in part by the difference in the central and western equatorial Pacific (Pizarro et al. the size of sea lion populations in each zone. For the area 2001), where the type of wind (eastern or western winds) from Arica to Antofagasta the population has been determines whether an equatorial KW will upwell or estimated at 59,657 individuals (Bartheld et al. 2008), while downwell the thermocline often tens of meters (Bertrand et in the area from Cobquecura to Mocha Island in the VIII al. 2008). The KW are coastally trapped by the force of the region, this value is only 8,189 individuals (Sepúlveda et rotation of the earth (Coriolis) along the continental al. 2011). In fact, in the former area there are colonies platform or shelf (Clarke 1983, 1992). During a warm period (Punta Lobos, Punta Piojo and Bandurrias del sur) with the KW are in downwelling conditions, while during a cold populations that exceed the total estimated population period the KW are in upwelling conditions (Chavez et al. for the entire VIII region (Bartheld et al. 2008). Another 2008). factor that can explain the observed differences is the The behavior of the co-variable MEI (one month fact that sea lion in the south-central area of Chile interact delayed) showed that the number of sea lions interacting mainly with coastal purse seine fleet of small pelagic, in with purse seine fishing operations tends to decrease when fact Muñoz et al. (2013) found Chilean herring the environment moves from cold (negative standardized (Strangomera bentincki) as the main food items of sea values) to warm condition (positive values). According to lions in this area. Barber & Chavez (1983), during warm condition the The result of the variable Distance to the closest colony anchovy tends to move to greater depths, below the warmer center is in agreement with the findings of other studies and less productive surface waters (Arntz & Fahrbach (Szteren & Páez 2002, Smith & Baird 2005, Sepúlveda et 1996), thus decreasing their availability to purse seine al. 2007). This inverse correlation is partly explained by fishing nets (Ñiquen & Bouchon 2002, Bertrand et al. 2004). This causes the fleet to alter the focus of the operations to the fact that the feeding areas of the sea lions are the jack mackerel, which is generally sought further from fundamentally shallow-coastal waters (Hevia 2013, the coast and the sea lion colonies. Hückstädt et al. 2014), and also the females need to remain close to their offspring (Smith & Baird 2005). The distance It has also been observed that during warm condition from the set to the colony center also interacted the abundance of the anchovy decreases and they move significantly with the variable Time of haul, which is an to greater depths, which means that the sea lions must indicator of sunlight intensity, showing an increased expend more energy per unit of prey consumed, which number of sea lions in sets conducted around midday would explain why the anchovy become less relevant to and at least 20 nm from the colony centers during breeding their diets during this period, where they are replaced by period. This effect disappears in non-reproductive period, demersal species (Soto et al. 2006). The variation in the where were observed similar levels in the abundance of availability of anchovy induced by warm condition effects, sea lions during sets that took place to less and more affects higher predators such as fishing vessels and sea than 20 nm from a colony center. This can be understood lions, which modify their target species in adverse as part of a strategy used by the animals to reduce heat conditions. Sets targeting anchovy went from 97% during stress during breeding periods by entering the sea before cold condition to 48% during warm condition. This causes the sun reaches its highest point in the sky (midday) and a decrease in the interaction with sea lions, though this returning to land in the late afternoon, thus decreasing can also be explained by the high significance of interaction their exposure to heat during the hours around midday between variables such as target species and MEI, or (Soto et al. 2006). According to Campagna & LeBoeuf between number of vessels and MEI, where in the latter (1988), this strategy is a behavioral adaptation seen in case the number of vessels has a clear mitigating effect on otariids living in tropical and subtropical areas, which are intraspecific competition of sea lions during cold condition, exposed to intense solar radiation and high air by decreasing the presence of the animals with the increase temperatures during the breeding period. This behavior in the number of vessels operating in the zone. This effect has been observed in females in Peru (Soto et al. 2006) disappears during warm condition due to the decrease in and young individuals in Chile (Sepúlveda et al. 2012). the number of sea lions interacting with the fishing operations.
486 González et al. South American sea lion and purse seine fishing in northern Chile The co-variable SST showed an inverse relation with would like to thank the anonymous reviewers for the number of sea lions indicating that the foraging providing critical suggestions resulting in significant activities of these marine mammals is related to coastal improvements of the manuscript. upwelling, more than to oceanic factors, which is explained by the distribution of their main prey, the LITERATURE CITED anchovy, a species that is associated with coastal ACAP. 2012. Minimum data requirements for monitoring upwelling (Swartzman et al. 2008). In fact, 84% of the seabird bycatch. Western and Central Pacific Fisheries catches of anchovy are obtained less than 20 nautical Commission, Kolonia WCPFC-SC8-2012/B-WP-07: 1-17. 7 miles from the coast (Böhm et al. 2012) .
7Böhm MG, C Hernández, G Perez, E Díaz, L Ossa, H Reyes, R Aravena, Z Young, F Cerna, V. Bocic, C Machuca, L Muñoz & M Troncoso. 2012. Programa de seguimiento del estado de situación de las principales pesquerías nacionales: Asesoría integral para la toma de decisiones en pesca y acuicultura, 2011, actividad 1: Recursos pelágicos Zona Norte, Informe Final, Convenio SUBPESCA-IFOP, Valparaíso, 265 pp.
Vol. 50, Nº 3, 2015 487 Revista de Biología Marina y Oceanografía Beverton R. 1985. Analysis of marine mammal-fisheries Hevia K. 2013. Áreas de alimentación del lobo marino común interaction. In: Beddington JR, RJH Beverton & DM Otaria flavescens en la zona norte de Chile. Tesis de Lavigne (eds). Marine mammals and fisheries, pp. 3-33. Ingeniería Ambiental, Facultad de Ciencias, Universidad de George Allen & Unwin, London. Valparaíso, Valparaíso, 93 pp. Bjørge A, T Bekkby, V Bakkestuen & E Framstad. 2002. Hückstädt LA & T Antezana. 2003. Behaviour of the southern Interactions between harbour seals, Phoca vitulina, and sea lion (Otaria flavescens) and consumption of the catch fisheries in complex coastal waters explored by combined during purse-seining for jack mackerel (Trachurus Geographic Information System (GIS) and energetics symmetricus) off central Chile. ICES Journal of Marine modelling. ICES Journal of Marine Science 59(1): 29-42. Science 60: 1003-1011. Bonner WN. 1982. Seals and man: A study of interactions, 170 Hückstädt LA, RA Quiñones, M Sepúlveda & DP Costa. pp. Washington Sea Grant Publication, University of 2014. Movement and diving patterns of juvenile male South Washington Press, Seattle. American sea lions off the coast of central Chile. Marine Burnham K & D Anderson. 2002. Model selection and Mammal Science 30: 1175-1183. multimodal inference: A practical information theoretic Königson SJ, KE Lündström, MMB Hemmingsson, SG approach, 488 pp. Springer-Verlag, New York. Lunneryd & H Westerberg. 2006. Feedings preferences Campagna C & BJ LeBoeuf. 1988. Thermoregulatory of harbour seals (Phoca vitulina) specialized in raiding behaviour of southern sea lions and its effect on mating fishing gear. Aquatic Mammals 32: 152-156. strategies. Behaviour 107: 73-90. Melvin EF, C Heinecken & TJ Guy. 2009. Optimizing tori Chavez FP, A Bertrand, R Guevara-Carrasco, P Soler & J line designs for pelagic tuna longline fisheries: South Africa. Csirke. 2008. The northern Humboldt Current system: Report of work under special permit from the Republic of brief history, present status and a view towards the future. South Africa. Department of Environmental Affairs and Progress in Oceanography 79: 95-105. Tourism, Marine and Coastal Management, Pelagic and High Seas Fishery Management Division.
488 González et al. South American sea lion and purse seine fishing in northern Chile Reyes P, R Hucke-Gaete & JP Torres-Florez. 2013. First Sielfeld W. 1999. Estado del conocimiento sobre conservación observations of operational interactions between bottom- y preservación de Otaria flavescens (Shaw, 1800) y trawling fisheries and South American sea lion, Otaria Arctocephalus australis (Zimmermann, 1783) en las costas flavescens in south-central Chile. Journal of the Marine de Chile. Estudios Oceanológicos 18: 81-96. Biological Association of the United Kingdom 93: 489- Smith MH & SJ Baird. 2005. Factors that may influence the 494. level of incidental mortality of Hooker’s sea lions Riedman M. 1990. The pinnipeds. Seals, sea lions, and (Phocarctos hookeri) in the squid (Nototodarus spp.) trawl walruses, 439 pp. University of California Press, Los fishery in SQU 6T. New Zealand Fisheries Assessment Angeles. Report 2005: 35-56. Sepúlveda M, MJ Pérez, D Oliva, LR Durán, W Sielfeld, V Soto KH, AW Trites & M Arias-Schreiber. 2006. Changes in Araos & M Buscaglia. 2007. Operational interaction diet and maternal attendance of South American sea lions between South American sea lions Otaria flavescens and indicate changes in the marine environment and prey artisanal (small-scale) fishing in Chile: results from abundance. Marine Ecology Progress Series 312: 277-290. interview surveys and on-board observations. Fisheries Swartzman G, A Bertrand, M Gutiérrez, S Bertrand & L Research 83: 332-340. Vasquez. 2008. The relationship of anchovy and sardine Sepúlveda M, D Oliva, A Urra, MJ Pérez-Álvarez, R to water masses in the Peruvian Humboldt Current System Moraga, D Schrader, P Inostroza, A Melo, H Díaz & W from 1983-2005. Progress in Oceanography 79: 228-237. Sielfeld. 2011. Distribution and abundance of the South Szteren D & E Páez. 2002. Predation by southern sea lions American sea lion Otaria flavescens (Carnivora: Otariidae) (Otaria flavescens) on artisanal fishing catches in Uruguay. along the central coast off Chile. Revista Chilena de Historia Marine and Freshwater Research 53: 1161-1167. Natural 84: 97-106. Venables WN & BD Ripley. 2002. Modern applied statistics Sepúlveda M, RA Quiñones, P Carrasco & MJ Pérez- with S, 504 pp. Springer, New York. Álvarez. 2012. Daily and seasonal variation in the haul-out Wickens P. 1995. A review of operational interactions between behavior of the South American sea lion. Mammalian Biology pinnipeds and fisheries. FAO Fisheries Technical Paper 77(4): 288-292. 346: 1-86. Sielfeld W, C Guerra, LR Durán, E Acuña, A Aguayo-Lobo, Wolter K & MS Timlin. 1998. Measuring the strength of M Sepúlveda, F Palma, A Malinarich, G Cerda, A ENSO-how does 1997/98 rank? Weather 53: 315-324. Bolvarán, R Grau, X Veloso, Y Guerra, M Vargas, N Amado, R Peredo & J Galáz. 1997. Monitoreo de la Wood SN. 2008. Fast stable direct fitting and smoothness pesquería y censo del lobo marino común en el litoral de la selection for generalized additive models. Journal of the I-IV Regiones. Informe Final Proyecto FIP 95-28: 1-105. Royal Statistical Society: Series B 70(3): 495-518. Subsecretaría de Pesca, Valparaíso.
Received 01 July 2014 and accepted 31 July 2015 Associate Editor: Maritza Sepúlveda M.
Vol. 50, Nº 3, 2015 489 Revista de Biología Marina y Oceanografía